LinearRegressionProcessor performs a linear regression on the hits in seed tracks. More...

#include <LinearRegressionProcessor.h>

Inheritance diagram for marlintpc::LinearRegressionProcessor:

Public Member Functions
	LinearRegressionProcessor ()
	Default constructor.

	~LinearRegressionProcessor ()
	Destructor.

Processor *	newProcessor ()
	Return a new instance of this processor.

void	init ()
	Called at the begin of the job before anything is read.

void	processRunHeader (lcio::LCRunHeader *run)
	Called once per run to process the event's header.

virtual void	processEvent (lcio::LCEvent *evt)
	Called for every event - the working horse.

virtual void	check (lcio::LCEvent *evt)
	Called for every event - right after processEvent() has been called for all processors.

virtual void	end ()
	Called after data processing for clean up in the inverse order of the init() method so that resources allocated in the first processor also will be available for all following processors.

Static Public Member Functions
static bool	compareHit (const std::pair< TrackerHit , double > &pair1, const std::pair< TrackerHit , double > &pair2)

Detailed Description

LinearRegressionProcessor performs a linear regression on the hits in seed tracks.

It calculates the track parameters for a straight line. The linear regression is an analytical calculation, but equivalent to a chi^2 fit. Does not need seed tracks, so it can directly work on track candidates. Orientation of the track can be either along x- or y-axis.

case1: $\displaystyle |\phi | < \frac{\pi}{4}$
y = ax + b
z = cx + d
$\displaystyle \phi = atan(a)$
$\displaystyle d_0 = b\cdot cos(\phi)$

calculation of $\lambda$ :
(angle betwenn track and x-y-plane, that is between track and projection on x-y-plane)
track: $\displaystyle \vec{g}(x) = \left(\begin{array}{c} x \\ ax + b \\ cx + d \end{array} \right)$
projection: $\displaystyle \vec{f}(x) = \left(\begin{array}{c} x \\ ax + b \\ 0 \end{array} \right)$
$\displaystyle cos( \lambda ) = \frac{(\vec{g}(1) - \vec{g}(0))\cdot(\vec{f}(1) - \vec{f}(0))}{|(\vec{g}(1) - \vec{g}(0))| \cdot |(\vec{f}(1) - \vec{f}(0))| }$
$\displaystyle \Rightarrow cos( \lambda ) = \frac{ \sqrt{ a^2 + 1 } }{ \sqrt{ a^2 + 1 + c^2 } }$
$\displaystyle z_0 = c\cdot x_0 + d = c\cdot sin( \phi )\cdot d_0$
case2: $\displaystyle |\phi | > \frac{\pi}{4}$
x = ay + b
z = cy + d
$\displaystyle \phi = \frac{\pi}{2}-atan(a)$
$\displaystyle d_0 = b\cdot sin(\phi)$
same as above:
$\displaystyle cos( \lambda ) = \frac{ \sqrt{ a^2 + 1 } }{ \sqrt{ a^2 + 1 + c^2 } }$
$\displaystyle z_0 = c\cdot y_0 + d = c\cdot sin( \frac{\pi}{2} - \phi)\cdot d_0$

dEdx is the charge deposition per track length. The procedure is the following:
1) one sorts the hits along the track (parametrization: $\displaystyle \vec{P}=\vec{d_0}+s\begin{array}{c} cos\phi \\sin\phi \end{array}$ $\Rightarrow$ sorting in s)
2) calculate distance from first to last point along track $\displaystyle l=|s_{last}-s_{first}|$
3) $\displaystyle l_{eff} = l\cdot\frac{n}{n-1}$ where n is the number of hits on the track 4) $\displaystyle dEdx = \frac{\sum Q}{l_{eff}}$
Author
M. Killenberg(University of Bonn), S.Zimmermann(University of Bonn)

Input - Prerequisites

A collection of track candidates or seen tracks (lcio::Tracks) containing the TrackerHits.
Note: The LCIO::TRBIT_HITS flag has to be set when this collection is created so the hits are stored in the track.

Output

Collection of straight tracks.

Parameters

InputTrackCandidates	The the name of the input collection of track candidates (default: TPCTrackCandidates)
OutputTracks	The name of the collection the straight tracks be stored under (default: TPCTracks)
SetOutputTransient	If not 0 the output collection is set transient (default: 0)

The documentation for this class was generated from the following files:

LinearRegressionProcessor.h
LinearRegressionProcessor.cc

Public Member Functions

Static Public Member Functions

Detailed Description

Input - Prerequisites

Output